The success of gene therapy techniques depends largely on the ability to achieve a combination of stable chromosomal integration and high-level, regulated expression of transferred genes in a manner safe to humans. Many current techniques allow efficient transient transfection of cells in vitro and in vivo with large DNA fragments. However, subsequent chromosomal integration is very inefficient. To overcome low levels of integration, retroviral vectors, which integrate very efficiently in permissive cells, can be used.
While recombinant retroviral vectors allow for integration of a transgene into a host cell genome, most retroviruses can only transduce dividing cells, which limits their use for in vivo gene transfer to nonproliferating cells such as hepatocytes, myofibers, hematopoietic stem cells, and neurons. Non-dividing cells are the predominant, long-lived cell type in the body, and account for most desirable targets of gene transfer, including liver, muscle, and brain. Even protocols attempting the transduction of hematopoietic stem cells require demanding ex vivo procedures for triggering cell division in these cells prior to infection.
One way of overcoming this obstacle is to employ lentiviral vectors, in place of conventional retroviral vectors. Lentiviruses are complex retroviruses which, based on their higher level of complexity, can integrate into the genome of nonproliferating cells and modulate their life cycles, as in the course of latent infection. These viruses include HIV-1, HIV-2 and SIV. Like other retroviruses, lentiviruses possess gag, pol and env genes which are flanked by two long terminal repeat (LTR) sequences. Each of these genes encodes multiple proteins, initially expressed as one precursor polyprotein. The gag gene encodes the internal structural (matrix capsid and nucleocapsid) proteins. The pol gene encodes the RNA-directed DNA polymerase (reverse transcriptase, integrase and protease). The env gene encodes viral envelope glycoproteins and additionally contains a cis-acting element (RRE) responsible for nuclear export of viral RNA. The 5′ and 3′ LTRs serve to promote transcription and polyadenylation of the virion RNAs. The LTR contains all other cis-acting sequences necessary for viral replication. Adjacent to the 5′ LTR are sequences necessary for reverse transcription of the genome (the tRNA primer binding site) and for efficient encapsidation of viral RNA into particles (the Psi site). If the sequences necessary for encapsidation (or packaging of retroviral RNA into infectious virions) are missing from the viral genome, the result is a cis defect which prevents encapsidation of genomic RNA. However, the resulting mutant is still capable of directing the synthesis of all virion proteins. A comprehensive review of lentiviruses, such as HIV, is provided, for example, in Field's Virology (Raven Publishers), eds. B. N. Fields et al., © 1996.
In addition to gag, pol, and env, lentiviruses, unlike other retroviruses, have several “accessory” genes with regulatory or structural function. Specifically, HIV-1 possesses at least six such genes, including Vif, Vpr, Tat, Rev, Vpu and Nef. The closely related HIV-2 does not code for Vpu, but codes for another unrelated protein, Vpx, not found in HIV-1.
The Vpr gene encodes a 14 kD protein (96 amino acids) (Myers et al. (1993) Human Retroviruses and AIDS, Los Alamos National Laboratory, N.M.). The Vpr open reading frame is also present in most HIV-2 and SIV isolates. Amino acid comparison between HIV-2 Vpr and Vpx shows regions of high homology suggesting that Vpx may have arisen by duplication of the Vpr gene (Myers et al. (1993), supra.). Vpr and Vpx are present in mature viral particles in multiple copies, and have been shown to bind to the p6 protein which is part of the gag-encoded precursor polyprotein involved in viral assembly (WO 96/07741; WO 96/32494). Thus, incorporation of Vpr and Vpx into viral particles occurs by way of interaction with p6 (Lavallee et al. (1994) J. Virol. 68: 1926-1934; and Wu et al. (1994) J. Virol. 68:6161). It has been further shown that Vpr associates, in particular, with the carboxy-terminal region of p6. The precise role of Vpr and Vpx is yet to be clearly determined, however the data to date suggests that these proteins have a role in the stage of early infection. It has also been shown that Vpr and Vpx, expressed in trans with respect to the HIV genome, can be used to target heterologous proteins to HIV virus (WO 96/07741; WO 96/32494). A description of the structure and function of Vpr and Vpx, including the full-length nucleotide and amino acid sequences of these proteins and their binding domains are also provided in WO 96/07741, as well as in Zhao et al. (1994) J. Biol. Chem. 269(22):1577 (Vpr); Mahalingham et al. 91995) Virology 207:297 (Vpr); and Hu et al. (1989) Virology 173:624) (Vpx). Other relevant references relating to Vpr include, for example, Kondo et al. (1995) J. Virol. 69:2759; Lavallee et al. (1994) J. Virol. 68:1926; and Levy et al. (1993) Cell 72:541. Other relevant references relating to Vpx include, for example, Wu et al. (1994) J. Virol. 68:6161. All of the aforementioned publication are incorporated by reference herein.
In view of the advantages associated with retroviral vectors in gene therapy, particularly lentiviruses which are capable of infecting non-dividing cells, improved methods for generating pure stocks of recombinant virus, free of replication-competent helper virus, would be of great value in the art. Recombinant retroviruses are generally produced by introducing a suitable proviral DNA vector into mammalian cells (“packaging cells”) that produce the necessary viral proteins for encapsidation of the desired recombinant RNA, but which lack the signal for packaging viral RNA (ψ sequence). Thus, while the required gag, pol, and env genes of the retrovirus are intact, there is no release of wild-type helper virus by these packaging lines. However, when the cells are transfected with a separate vector containing the ψ sequence required for packaging, wild-type retrovirus can arise by recombination (Mann et al. (1983) Cell 33:153). This is a major danger, particularly in the case of lentiviruses, such as HIV.
Current approaches to avoid the safety dangers associated with recombination leading to production of replication-competent helper virus include making additional mutations (e.g., LTR deletions) in the viral constructs used to create packaging lines, and separating the viral genes necessary for producing virions onto separate plasmids. For example, it has recently been shown that recombinant Moloney murine leukemia virus (MuLV), free of detectable helper-virus, can be produced by separating the gag and pol genes from the env gene in packaging cells (Markowitz et al. (1998) J. Virol. 62(4):1120). These packaging cells contained two separate plasmids collectively encoding the viral proteins necessary for virion production, reducing the likelihood that the recombination events necessary to produce intact retrovirus (i.e., between three plasmid vectors) would occur when cotransfected with a third vector containing the ψ packaging signal.
Additional methods for producing safer retroviral packaging cell lines, particularly lentiviral packaging cell lines, which generate recombinant retrovirus, yet do not themselves either yield detectable helper virus or transfer viral genes, would be of great value in human gene therapy.